Supplementary Materials for
A stretchable wireless wearable bioelectronic system for multiplexed
monitoring and combination treatment of infected chronic wounds
Ehsan Shirzaei Sani
et al.
Corresponding author: Wei Gao, weigao@caltech.edu
Sci. Adv.
9
, eadf7388 (2023)
DOI: 10.1126/sciadv.adf7388
The PDF file includes:
Materials and Methods
Notes S1 and S2
Figs. S1 to S32
Legend for movie S1
References
Other Supplementary Material for this manuscript includes the following:
Movie S1
Note S1. Optimization of enzymatic sensors for wound fluid analysis
The past decades have witnessed tremendous success in developing enzymatic sensors (e.g.,
continuous glucose monitors) for continuous monitoring of circulating metabolites in blood,
interstitial fluid (ISF), or non
-
invasively accessible alternative fluids
(e.g., sweat, saliva, and tears).
However, complex and heterogeneous composition of wound fluid (e.g. high protein levels, local
and migrated cells, and exogenous factors such as bacteria) leads to severe and unique matrix
effects for previously reported e
nzymatic sensors and failure in accurate measurement of the target
metabolite levels in untreated wound fluid. There are few reports on biosensors that are able to
perform continuous long
-
term wound fluid metabolic analysis
(
31
)
.
Taking glucose sensors as the exemplar, we examined a number of glucose sensor configurations
as demonstrated in
figs. S8
and
S9
. Instable or non
-
responsive glucose sensor responses were
commonly observed in simulated wound fluid (SWF) samples or wound flu
id (WF) samples
collected from diabetic rats. This indicates that previously reported glucose sensors suffer from
severe matrix effects and fail to accurately measure the target metabolite levels in untreated WF.
Moreover, high levels of metabolites in dia
betic wound fluid, particularly glucose (up to 50 mM),
pose another major challenge to obtain linear sensor response in the physiological concentration
ranges.
Specifically, various glucose sensors with Prussian blue (PB) as the redox mediator were prepare
d
based on the reported methods. Glucose sensors were first prepared by modifying glucose oxidase
(GOx), GOx/bovine serum albumin (BSA), GOx/polyaniline (PANI), and GOx/chitosan (CS) on
the top of the Au/PB electrodes. The first three types showed a poor l
inear relationship between
amperometric responses and the physiological diabetic glucose concentration ranges (up to 50
mM) in SWF. In contrast, the GOx/CS/MWCNTs
-
based sensor showed good linear response,
potentially due to the limited glucose diffusion in
the CS matrix. All these sensors did not respond
to the addition of glucose in WF, indicating the severe matrix influences of complex wound fluid
to enzymatic sensor performance.
In order to increase sensor range and minimize biofouling effects, we explor
ed the use of an outer
porous membrane that serves as a diffusion limiting layer to protect the enzyme, tune response,
increase operational stability, as well as enhance the linearity and sensitivity magnitude of the
sensor. We next fabricated our enzymati
c GOx/CS/MWCNTs glucose sensor with additional
porous membrane coatings including CS, poly(ethylene glycol) diglycidyl ether (PEGDGE),
Nafion, and polyurethane (PU) (
fig. S9
). As expected, the addition of diffusion layers indeed
improves the sensor’s linea
r range in SWF. However, CS, PEGDGE, and Nafion coated sensors
did not show good response in wound fluid upon the addition of glucose. We observed that PU
-
based sensor showed the highest linearity over the wide physiological concentration range as well
as
high reproducibility in complex wound fluid matrix (
fig. S10
).
The optimized enzymatic sensors with PU coating possess several critical design features to
operate in the wound bed environment for a prolonged period. First, enzymes such as glucose
oxidase
are specific for the target analyte in the wound fluid, which minimize the interference of
the other biochemicals including electrolytes and metabolites. Second, the CS/CNT layer allows
efficient electron transfer but reduces electrode poisoning due to int
erferences from undesired
exogenous biochemicals in the wound fluid. Third, the PU layer improves long
-
term stability of
sensor for accurate and stable
in vivo
function. In addition, polyurethane’s biocompatibility
improves
in vivo
durability of the sensor
and eliminates safety concerns. Moreover, the PU mass
transport limiting membrane has excellent mechanical strength which improves the sensor physical
stability.
Note S2. Analysis of gene expression in the wound healing process
During wound healing, the
extracellular matrix (ECM) at the wound site undergoes dramatic
reorganization. An elevated expression of collagen type I alpha 1 (Col1a1) and collagen type III
alpha 1 (Col3a1) was observed in the electrical stimulation (ES) and combination therapy groups
as compared to control group on day 3 (
fig. S31A
and
B)
. This can be mainly attributed to
fibroblasts proliferation. In this process, substantial quantities of matrix proteins (predominantly
collagen types I and III) are synthesized and deposited, resulti
ng in improved tensile strength of
the regenerated wound skin. Interestingly, we also observed a substantial increase in Col3a1
expression in the control group as compared to other groups on day 14. This is potentially due to
downregulation of matrix metal
loproteinase
-
9 (MMP
-
9) that resulted in ECM accumulation during
the wound healing process and yielded to keloid or hypertrophic scarring in the control group
(
64
)
.
MMPs are
calcium
-
dependent zinc
-
containing endopeptidases that collectively degrade and resorb
all major components of the ECM
(
65
)
. The upregulation of MMPs during the wound repair
p
rocess contributed to scarless healing in the combination therapy group after 14 days (
fig. S31C)
.
Higher expression levels of fibroblast growth factor 10 (Fgf10), C
-
X
-
C motif chemokine ligand 1
(Cxcl1), and vascular endothelial growth factor A (Vegfa) we
re observed for the combination
therapy group when compared to other groups (
fig. S31D
and
F)
. Such higher expression is
associated with early angiogenesis and neovasculature formation during chronic wound healing
(
66
)
. Angiogenesis is the process of newly formed blood vessels and plays a crucial role in
supplying necessary nutrien
ts to the new granulation tissue. Angiogenesis and neovascularization
in the new ECM are triggered by several growth factors including Fgf2 and Fgf10
(
66
)
. Fgfs
generally regulate angiogenesis
via
the recruitment of inflammatory cells that results in up
-
regulation of various chemokines such as chemokine (C
-
C motif) ligand 2 (Ccl2)
and Cxcl1
(
67
)
.
During inflammation, Cxcl chemokines regulate the timely recruitment of specific populations of
leukocytes to the damage site. They are also important
in angiogenesis, tumor formation, and
tumor metastasis
(
68
)
. Vegfa is also critical for enhancing angiogenesis in the early stages of
wound healing, particularly by promoting endothelial cell
s migration
(
69
)
. A significant
expression of Mito
gen
-
activated protein kinase 1/2 (Mapk1/2) was also observed in the
combination therapy group as compared to other groups (
fig. S31G)
. Generally, Vegfa plays a key
role in multiple endothelial cell
-
specific functions including activation of the downstream
Mapk1/2 pathway, therefore promoting cell migration, proliferation, and angiogenesis.
The higher expression of Pten gene also confirms the
immunohistochemistry
results that the
positive effect of electrical stimulation on preferential activation of
voltage
-
gated channels
facilitated cell migration and orientation (electrotaxis) (
Fig. 6E
and
fig. S31E,H)
(
42
)
.
Collectively, these findings confirmed that, through combination therapy,
the wearable patch could
modulate cell proliferation, migration, and ECM deposition and remodeling, enabling an
accelerated scarless cutaneous wound healing.
Fig. S
1
.
The
f
abrication process of
the
stretchable
wearable patch
.
Fig. S1.
Fabrication process of the patch
Check the printability
of the hydrogel (if
needs rheological
properties)
Cleaning wafer
Etching Cu & transferring pattern
Depositing Cu
C
oating parylene
Patterning Au/Cr
Etching parylene
Coating SEBS
Laser
-cutting patch
Fig. S
2.
Optical i
mages of the
stretchable
wearable patch
.
(
A
and
B
)
Photograph of
stretchable
and flexible
wearable patche
s
.
Scale bars, 1 cm.
Fig. S2. Pictures of fabrication
device
Scale bar
A b 1 cm
Scale bar
C 100 um
A
B
C
D